Relieving fuel poverty in Wales with external wall insulation

Jo Atkinson BSc, PhD Academic Associate, Ecological Built Environment Research and Enterprise, Cardiff School of Art and Design, Cardiff Metropolitan University, Cardiff, UK John Littlewood BSc, PhD, FHEA, ACIAT Head of EBERE and Senior Lecturer, Ecological Built Environment Research and Enterprise (EBERE), Cardiff School of Art and Design, Cardiff Metropolitan University, Cardiff, UK (corresponding author: jlittlewood@cardiffmet.ac.uk)


Introduction
In 2009 the Welsh government introduced the 'Arbed' (meaning 'save' in Welsh) scheme as part of its policy agenda to improve the energy efficiency of existing dwellings in the top 10% most deprived areas in Wales (Welsh Assembly Government, 2010a).In addition to reducing carbon dioxide emissions, the policy agenda was focused on reducing levels of fuel poverty for the occupants living in these dwellings (Welsh Assembly Government, 2010a).To demonstrate the effectiveness of the scheme and thus the successfulness of the policy, monitoring and evaluation was a funding requirement of the Arbed scheme.However, there was no methodology and very little funding for these assessments as part of the first phase of the Arbed scheme (Arbed I hereafter).
In collaboration with two of the housing associations that successfully won their bids for funding through Arbed I, a doctoral research project was developed and implemented to undertake the assessments (which was successfully completed by the first author in 2015).As the doctoral research project progressed, a large gap was identified in the empirical baseline data for assessing the effect that retrofitted energy efficiency interventions has on reducing energy consumption and carbon dioxide emissions (May, 2012;May and Rye, 2012;Rye and Hubbard, 2012).One of the key reasons for the gap is due to the over-reliance on modelled and theoretical data to undertake these types of assessments in the UK (Fitton and Brown, 2015;May and Rye, 2012).This paper documents the results of energy performance assessments undertaken to establish the effectiveness of retrofitted external wall insulation (EWI) at five pre-1919 case study dwellings in Swansea, Wales.Retrofitting EWI was the predominant intervention employed by both housing associations as part of their Arbed I schemes, due to the quantity and the existing energy performance of many of their pre-1919 dwellings.The data collected to undertake these assessments consisted of energy consumption and costs from utility bills, carbon dioxide emissions, floor area, heating degree days and thermal comfort perceptions using occupant surveys (Hopper et al., 2012a).In addition, cost data were collected to determine the payback of the interventions.Furthermore, this paper discusses the unintended consequences of retrofitting the EWI, as well as the lessons learnt from Arbed I that can be carried forward for developing further policies and predicting outcomes from retrofit works to dwellings.This includes the first author implementing the lessons learnt from the doctoral research project by providing technical support to scheme managers currently managing the latest round of Arbed funding in 2015/2016, as part of her role working for the Carbon Trust in Wales.The schemes are being delivered across Wales and include hard-to-treat cavity wall and system build dwellings, as well as the pre-1919 dwellings that provided the case studies for the doctoral research project.

The policy context for retrofitting EWI on existing dwellings in Wales
In the UK the overarching drivers for improving the energy performance of existing dwellings are the legally binding targets set out in the Climate Change Act 2008Act (2008)), which states that carbon dioxide emissions need to be reduced by 80% by 2050 using the 1990 levels as the baseline.However, it is anticipated that over two thirds of existing dwellings will still be in use in 2050 (Department for Communities and Local Government, 2008).Furthermore, Wales has some of the oldest and poorest thermal performing dwellings in Europe, with approximately 34% having been built before 1919 (BPIE, 2011;Gleeson et al., 2011;King, 2011;Welsh Assembly Government, 2005).
In addition to poor thermal performance, the occupants that live in these dwellings are at increased risk of having to live in fuel poverty, as well as suffering from poor health, which could lead to an increased demand on the National Health Service (NHS) (Davidson et al., 2011;Howarth, 2010;Nicol et al., 2010;Thompson and Mitchell, 2015).
In recognition of the scale and urgency of improving the energy performance of existing dwellings in Wales, the Welsh government set up the Arbed scheme in 2009 and work started in 2010 (Welsh Assembly Government, 2010a).To complement the UK targets, the Welsh government set their own target for reducing carbon dioxide emissions from sectors where they had devolved responsibility as part of their policy framework for combating climate change (Welsh Assembly Government, 2010b).The target set is for an annual 3% reduction of carbon dioxide emissions from 2011, relative to a baseline of average emissions between 2006 and 2010.It was envisaged that the Arbed scheme would make a contribution from the domestic sector (Welsh Assembly Government, 2010b).Arbed is in addition to any other scheme such as the UK-wide Energy Company Obligation (ECO), which is the UK government strategy for improving the energy efficiency of the most inefficient existing dwellings in the UK (Ofgem, 2015).

Data collection
As set out earlier, the purpose of this paper is to discuss the actual post-retrofit energy performance of five Arbed I case study dwellings relative to a pre-retrofit baseline; these results are then compared with the original intent of the Arbed I scheme.To establish the pre-retrofit and post-retrofit energy performances, quantitative data were collected using established building performance evaluation methods.These methods included energy consumption monitoring involving energy efficiency calculations, which are based on pre-retrofit and post-retrofit energy usage, carbon dioxide emissions and costs, as well as dwelling floor area and heating degree days, and occupant surveys to ascertain perceptions about thermal comfort and behaviour since having the EWI was installed and thus identify any potential relationships to minimal or non-existent reductions in energy consumption.The authors would have liked to monitor pre-retrofit and post-retrofit internal temperatures compared with external temperatures and other climatic information.However, it was not possible to fund the sensors and data loggers to record and log these temperature data, as the grant did not stretch to cover the total cost of these items in addition to other equipment already purchased.
Combined with the energy usage, carbon dioxide (CO 2 ) emissions, costs and dwelling floor area data, the heating degree days allow the resulting data to be normalised to take account of the differences between the external weather conditions of two heating seasons and, thus, for the results to be presented in kilowatt-hours per square meter per year, CO 2 equivalent per square metre per year and pounds per square metre per year.
To collect the energy usage and cost data, each case study dwelling's energy supplier was contacted directly for the information.Consent to collect these data was obtained during the post-retrofit occupant surveys.The rationale for this approach was due to discovering the lack of availability of these vital data during the pre-retrofit occupant surveys.For the majority of the case study dwellings, the occupants either were on a pre-payment meter for their electricity and gas or did not keep their energy bills after they had paid them.Once the energy consumption data had been collected, the corresponding greenhouse emissions were calculated using the conversion factors supplied by the Carbon Trust (Carbon Trust, 2013).For electricity the conversion factor was 0•44548 per kilowatt-hour, and for mains gas the conversion factor was 0•18404 per kilowatt-hour in 2013 (Carbon Trust, 2013).
Where the floor area data were not held by the housing associations on their databases, these were collected as part of the post-retrofit occupant surveys.The heating degree days' data were collected from a local weather station.To normalise the energy data using the heating degree days, the cumulative energy consumption for the 12 months prior to the EWI being installed was divided by the number of heating degree days for the same 12 months to ascertain the kilowatt-hours per heating degree day; this was then multiplied by the 20-year average annual heating degree day to establish the normalised annual energy consumption.This calculation is illustrated in the following equation (Atkinson, 2015) Normalised energy consumption ðkWhÞ ¼ annual energy consumption ðkWhÞ annual heating degree days Â 20-year average heating degree days 1.
The calculation from Equation 1 was then repeated with postretrofit data.no savings had been made.Energy costs were also normalised relative to the weather, to allow a direct comparison between preretrofit and post retrofit.These results were then analysed in conjunction with the occupant's responses to questions about their perceived comfort levels after the EWI was retrofitted compared with that before.To collect these perceived thermal comfort data as part of the occupant surveys (structured interviews), the two main questions asked were 'Do you feel that your home is warmer since it has been insulated?' and 'Have you increased the internal temperature that you keep your home at during the heating season (winter) since having the external wall insulation installed?' Occupants were given three choices to answer both questions: 'yes', 'no' or 'don't know'.However, for the second question, there were two sub-questions which followed on from the answer.For these sub-questions, the occupants were given the option to select more than one response.

The effect of retrofitted EWI on energy use, costs and carbon dioxide emissions
Focusing on the results from five mid-terrace case study dwellings, this section sets out the occupants' responses to the questions about thermal comfort and behaviour, along with the energy consumption and cost data collected from the corresponding energy company.Together with the calculated carbon dioxide emissions, the energy consumption and cost data have been normalised to take account of the external weather conditions for the two heating seasons before and after the EWI was installed using the formulas set out above from local heating degree days; these results are then divided by the floor area to give the final units.These five dwellings were chosen for this paper as they are all mid-terrace case studies from the doctorate study and this is the only dwelling type for which construction cost data were also available from the housing associations.For the doctorate, there is energy consumption and carbon dioxide emissions data for 12 dwellings and energy cost data for 10 dwellings, which include end-terrace houses and flats.
Collectively these data support the inferences that have been drawn from these five case studies set out in this paper.The authors acknowledge that five dwellings represent 42% of the total sample dwellings included within this doctoral study.

Case studies
Each of the five case studies are a mid-terrace and either an owner-occupied dwelling or a dwelling owned by one of the housing association partners with tenants which received retrofitted EWI in the spring of 2011.The energy consumption and heating degree day data are presented for the 12 months before and after this time.It should be noted that during the data collection period other than the EWI upgrade, the other fabric and energy systems remained unchanged, to the researcher's knowledge.

Case study I
Case study I has an internal floor area of 77 m 2 , and the EWI was retrofitted in June 2011.At the time of both the pre-retrofit and post-retrofit occupant surveys (February 2011 and September 2012 respectively), there were two occupants between 18 and 65 years.In a typical week at the pre-retrofit stage, the dwelling was occupied in the afternoons and evenings and at night throughout weekdays and continuously at weekends.At the post-retrofit stage, the dwelling was continuously occupied during weekdays and weekends in a typical week.
As set out in Table 1, the resulting pre-retrofit and post-retrofit energy consumption and carbon dioxide emissions indicate that there has been a 13% overall increase, which equates to 983 kWh and 438 kg of carbon dioxide equivalent per year.With reference to thermal comfort perceptions and behaviour responses in the post-retrofit occupant survey for this particular dwelling, the occupants stated that they felt their home was warmer since having the EWI installed and they did not know whether they had increased the internal temperature inside their home.The increase in energy consumption could be attributed to an increase in the internal temperature.However, the increase could also be attributed to the change in occupancy pattern or a combination of the two reasons.
Due to a 4% increase in energy costs between the pre-retrofit and post-retrofit stages, the figures in Table 2 indicates that the occupants are even more unlikely to realise any savings on their electricity bill; this is based on the normalised energy costs, which indicates an overall approximate 17% increase.Furthermore, the following year (2012), the occupants' electricity costs went up by a further 7%.As a result, the retrofitted EWI is unlikely to assist with alleviating the occupants out of fuel poverty; however, the severity is likely to have been reduced.In addition, the results indicate that the occupants are unlikely to be better off financially.

Case study II
Case study II has an internal floor area of 104 m 2 , and the EWI was retrofitted in April 2011.At the time of both the pre-retrofit and post-retrofit occupant surveys (February 2011 and September 2012 respectively), there was one occupant over the age of 65 years.In a typical week at both the pre-retrofit and post-retrofit stages, the dwelling was continuously occupied during weekdays and weekends.
As set out in Table 3, the resulting pre-retrofit and post-retrofit energy consumption and carbon dioxide emissions indicate that there has been a 23•27% overall reduction, which equates to 2754 kWh and 507 kg of carbon dioxide equivalent per year.With reference to thermal comfort perceptions and behaviour responses in the post-retrofit occupant survey for this particular dwelling, the occupant stated that they did not know whether their home was warmer since having the EWI installed and they had not increased the internal temperature inside their home.The results set out in Table 3 appear to support the occupant's statement that they had not increased the internal temperature in their home since having the EWI installed.
Due to a 28% increase in energy costs between the pre-retrofit and post-retrofit stages, the figures in Table 4 indicate that the occupant is likely to realise an approximate overall 5-7% increase in their gas bill; this is based on the normalised energy costs.Nevertheless, this is 23% less than it would otherwise had been if the EWI had not been retrofitted.The following year (2012), this occupant's gas costs went down by 7%.Therefore, it is likely that the occupant would have returned to the same level of energy costs as before the EWI was installed or potentially realised an approximate saving of 2%.As a result, the retrofitted EWI is unlikely to have assisted with alleviating this occupant out of fuel poverty.However, the severity is likely to have been reduced.

Case study III
Case study III has an internal floor area of 75 m 2 , and the EWI was retrofitted in June 2011.At the time of both the pre-retrofit and post-retrofit occupant surveys (February 2011 and September 2012 respectively), there was one occupant aged between 18 and 65 years.At the pre-retrofit stage, the dwelling was continuously occupied during a typical week.At the post-retrofit stage, the dwelling was typically occupied during weekday mornings and evenings and at night and during afternoons and evenings and at night at weekends.
As set out in Table 5, the resulting pre-retrofit and post-retrofit energy consumption and carbon dioxide emissions indicate that there has been a 5•45% overall reduction, which equates to 698 kWh and 128 kg of carbon dioxide equivalent per year.With reference to thermal comfort perceptions and behaviour responses in the post-retrofit occupant survey for this particular dwelling, the occupant stated that they felt that their home was warmer since having the EWI installed and they had not increased the internal temperature inside their home.The results set out in Table 5 appear to support the occupant's statement that they had not increased the internal temperature in their home since having the EWI installed.However, it is recognised that some of the reductions in energy consumption could also be attributed to the change in occupancy pattern.
Due to an 18% increase in energy costs between the pre-retrofit and post-retrofit stages, the figures in  occupant is likely to realise an approximate overall 13% increase in their gas bill; this is based on the normalised energy costs.Nevertheless, this is 5% less than it would otherwise had been if the EWI had not been retrofitted.The following year (2012), this occupant's gas costs went up by a further 1%.As a result, it appears that the retrofitted EWI is unlikely to assist with alleviating this occupant out of fuel poverty; however, the severity is likely to have been reduced.The results indicate that the occupants are more likely to be in deeper fuel poverty.

Case study IV
Case study IV is a mid-terrace house, which is rented from one of the housing associations and has an internal floor area of 110 m 2 .The EWI was retrofitted in June 2011.Both a pre-retrofit and a post-retrofit occupant survey were undertaken at this dwelling using the survey method for interviews.At the time of both the pre-retrofit and post-retrofit occupant surveys (March 2011 and October 2012 respectively), there were four occupants, two aged between 5 and 17 years and two aged between 18 and 65 years.At the pre-retrofit stage, the dwelling was continuously occupied during a typical week.At the post-retrofit stage, the dwelling was typically occupied during weekday mornings and evenings and at night and continuously on weekends.
As set out in Table 7, the resulting pre-retrofit and post-retrofit energy consumption and carbon dioxide emissions indicate that there has been a 5•27% overall increase, which equates to 1312 kWh and 242 kg of carbon dioxide equivalent per year.With reference to thermal comfort perceptions and behaviour responses in the postretrofit occupant survey for this particular dwelling, the occupant stated that they felt that their home was warmer since having the EWI installed and they had not increased the internal temperature inside their home.However, the results of the energy consumption monitoring set out in Table 7 do not correspond with the occupant's responses.
Due to a 16% increase in energy costs between the pre-retrofit and post-retrofit stages, the figures in Table 8 indicate that occupants are even more unlikely to realise any savings on their gas bill; this is based on the normalised energy costs, which indicate an overall approximate 20-21% increase.Furthermore, the following year (2012), the occupants' gas costs went up by a further 9%.Therefore, it is likely that the occupants would have realised a total increase of 29-30% on their energy costs after the latter energy cost change.As a result, the retrofitted EWI is very unlikely to assist with alleviating the occupants out of fuel poverty.The results indicate that the occupants are more likely to be in deeper fuel poverty.

Case study V
Case study V is a mid-terrace house which is rented from one of the housing associations and has an internal floor area of 100 m 2 .The EWI was retrofitted in June 2011.Both a pre-retrofit and a post-retrofit occupant survey were undertaken at this dwelling using structured interviews.At the time of both the pre-retrofit and post-retrofit occupant surveys (May 2011 and October 2012 respectively), there were four occupants; two aged between 5 and 17 years and two aged between 18 and 65 years.At both the pre-retrofit and post-retrofit stages, the dwelling was continuously occupied during a typical week.

Post
As set out in Table 9, the resulting pre-retrofit and post-retrofit energy consumption and carbon dioxide emissions indicate that there has been a 21•04% overall reduction, which equates to 6875 kWh and 1331 kg of carbon dioxide equivalent per year.With reference to thermal comfort perceptions and behaviour responses in the post-retrofit occupant survey for this particular dwelling, the occupants stated that they felt that their home was not warmer since having the EWI installed and they had not increased the internal temperature inside their home.The results set out in Table 9 appear to support the occupants' statement that they had not increased the internal temperature in their home since having the EWI installed.Nevertheless, it is recognised that in addition to the retrofitted EWI, there could be other explanations for the reductions in energy consumption.
Due to a 25% increase in energy costs between the pre-retrofit and post-retrofit stages, the figures in Table 10 indicate that the occupants are likely to realise an approximate overall 4% to 5% increase on their gas bill; this is based on the normalised energy costs.Nevertheless, this is 20-21% less than it would otherwise had been if the EWI had not been retrofitted.The following year (2012), the occupants' gas costs went down by 2%.Therefore, it is likely that the occupants would have realised only a 2-3% overall increase on their gas bill.As a result, the retrofitted EWI is unlikely to have assisted with alleviating the occupants out of fuel poverty.

EWI costs
Only limited capital cost data for retrofitting the EWI were available from one of the housing associations.The capital cost data that were provided were for two types of mid-terrace houses (with and without a rear annex) and included all additional works and preliminaries.The additional works included replacing fascias, gutters, downpipes and window sills, as well as temporarily removing and re-fixing fixtures, such as satellite dishes, aerials, washing lines and outside taps.The preliminaries included scaffolding, skips, personal protective equipment, sanitary facilities and labour.The breakdown for the two types of mid-terrace dwellings is set out in Table 11.However, it should be noted that these costs are based on 40 dwellings (20 of each type) being retrofitted as part of a whole-street approach, and, therefore, the costs for the preliminaries are not representative of a single dwelling installation.Nevertheless, these data are representative of three of the five Arbed I case study dwellings discussed in this paper.
Case studies I-III are mid-terrace dwellings with annexes at the rear and were part of a whole-street approach.However, due to there being no energy cost savings at any of these case study dwellings, it can be determined without any further analysis that the retrofitted EWI will not pay for itself with the energy prices illustrated in this paper.Therefore, to establish if this lack of payback is due to the exceptionally high energy price increase during the 12 months after the EWI was installed, a basic cost analysis has been undertaken which is solely based on the    occupant's pre-retrofit energy costs at the three case study dwellings.This basic analysis is set out in Table 12.From the analysis, it can be determined that it will take case study II approximately 89 years for the EWI to pay for itself and 414 years for case study III.As case study I increased their energy consumption, there is no payback for the EWI even where only pre-retrofit energy costs are used.It should be noted that this basic cost analysis does not take into account any other current or future energy price increases.Nevertheless, while it could be determined that case studies II and III will maintain 23% and 5% savings relative to the current and future energy costs, as there was no immediate saving, the payback will never be realised.

Post
Figure 1 illustrates the overall energy cost percentage changes relative to energy consumption changes resulting from the retrofitted EWI, which takes into account energy price changes that occurred for each dwelling between the pre-retrofit and post-retrofit stages.

Discussion
First and foremost the results indicate that the main objectives of the Arbed scheme, which were to contribute to Wales' policy agenda of reducing carbon dioxide emissions and fuel poverty, were only partly achieved (Welsh Assembly Government, 2010a).While some carbon dioxide emission savings were achieved, none of the occupants were taken out of fuel poverty, although the EWI will have helped with reducing the severity of the increased energy costs.In addition, due to the very limited reduction in energy consumption, which occurred only at the two dwellings, coupled with increases in energy costs, the payback for the retrofitted EWI was beyond its life expectancy, if achieved at all.In terms of payback, energy price rises actually helps shorten the payback period.For example, if mains natural gas was suddenly £1000/kWh, the EWI would be paid for in a week.
The rationale for not achieving any alleviation from fuel poverty can be attributed to the increase in energy costs, as well as the rebound effect.The occurrence of the rebound effect was expected and is consistent with findings from studies undertaken by other researchers, such as Boardman (2007), Herring and Roy (2007), Hong et al. (2009) and Palmer and Cooper (2011).Occupants at three of the five case study dwellings stated that they felt their home was warmer since the EWI was installed; one stated that they were not sure, and one said that it was not warmer.However, where increases in thermal comfort have occurred, these are likely to lead to reductions in demand on the UK's NHS, which could lead to indirect benefits to the society as a whole (Davidson et al., 2011;Nicol et al., 2010).In the opinion of the authors, these unintended benefits could outweigh these issues as long as they are anticipated and, thus, energy and carbon dioxide emission savings are not the primary driver for implementing retrofitted EWI.As a result, these benefits will therefore be anticipated and thus 'sold' as part of the outcome of future EWI retrofits, such as those funded through future Arbed schemes and ECO.
Despite all of the above, funding initiatives and grant schemes such as Arbed is, in the authors' opinions, a good way (and probably the most appropriate way) to fund retrofitted EWI in the domestic sector.This is due to Arbed funding coming directly from the Welsh government and, therefore, costs were not passed onto the customers.The occupants were not required to make a financial contribution in order for their homes to be upgraded with EWI.Thus, even with lengthy payback periods, considerably beyond lifetimes in some of the case studies, this is why Arbed is more appropriate than improvements through energy companies, who require a cost input from the occupants.For example, if EWI is installed through ECO, then not realising the energy savings will mean that the energy companies will not meet their carbon dioxide emission reduction targets.Not meeting these targets results in penalties, usually financial, being issued to the energy company.The energy companies will then need to recoup this money, and the most likely way that they will do this is to increase the cost of energy for the consumer, and for those households already in fuel poverty, this will only make the financial situation worse.In addition, Arbed I makes some improvement in occupant comfort (Atkinson, 2015).2008, 2008;Welsh Assembly Government, 2010a).Thus, in collaboration with one of the housing associations, Cardiff Metropolitan University is seeking ways in which to obtain data from all households which had Arbed I funding through their organisation, for EWI to establish whether the problems identified in this paper occur in these properties also; this work is ongoing in 2016.Furthermore, it is hoped that the results of this paper and other publications by the authors will assist the Welsh government to fund and collaborate with further studies to evaluate the whole Arbed programme, in the context of their own carbon dioxide emission targets and the Climate Change Act 2008 (Atkinson et al., 2015;Hopper et al., 2012aHopper et al., , 2012b)).For example, since completing her PhD in 2015, the first author is implementing the lessons learnt from the doctoral research project by providing technical support to domestic scheme managers currently managing the latest round of Arbed funding in 2015/ 2016, as part of her role working for the Carbon Trust in Wales.

Conclusions
This paper has presented findings from a doctoral research project monitoring the effectiveness of retrofitted EWI to dwellings in deprived communities in Wales as part of the Arbed I scheme.More specifically, the paper presents empirical baseline data for assessing the effect that the retrofitted interventions has on reducing energy consumption and carbon dioxide emissions, as well as alleviating occupant fuel poverty.This work is in collaboration with two social housing providers in Swansea, UK, where EWI was the predominant intervention, due to the quantity and existing energy performance of many of their pre-1919 dwellings.From the analysed data on pre-retrofit and post-retrofit energy consumption and carbon dioxide emissions, as well as cost data to determine the payback of the intervention, it has been demonstrated that, while some carbon dioxide emission savings were achieved, none of the occupants were taken out of fuel poverty.There was a very limited reduction in energy consumption, coupled with increases in energy costs, and, thus, the payback for the retrofitted EWI was beyond its life expectancy, if achieved at all.
Occupants at three of the five case study dwellings stated that they felt their home was warmer since the EWI was installed, and these increases in thermal comfort are likely to lead to reductions in demand on the UK's NHS, which could lead to indirect benefits to the society as a whole.These unintended benefits could outweigh the issues, as long as they are anticipated, and, thus, energy and carbon dioxide emission savings are not the primary driver for the retrofitted intervention.As a result, these benefits will therefore be anticipated and thus sold as part of the outcome of future EWI retrofits.The results of this research have demonstrated that Arbed is a really good method of funding retrofitted EWI.However, it is unlikely that retrofitted EWI will contribute to meeting the requirements of the Climate Change Act 2008 and the Welsh government's own carbon dioxide emission reduction targets.
In order to further validate the results presented in this paper, the authors note that monitoring needs be widened to include both temperature and relative humidity, both internally and extremely; however, due to budget restrictions for this project, this was not possible.
Finally, following the findings presented in this paper, the authors hope that it will assist the Welsh government to fund and collaborate with further studies to evaluate the whole Arbed programme, in relation to their 3% annual target to cut carbon dioxide emissions and whether it delivers the needs of the UK Climate Change Act.
The difference was determined by deducting the normalised pre-retrofit energy consumption from the normalised post-retrofit energy consumption.Where the result was a negative figure, this demonstrated that energy consumption had reduced.However, where the result was a positive figure, this denoted that

Table 2 .
Table 6 indicate that the Energy cost data for case study I

Table 3 .
Energy consumption and carbon dioxide emission data for case study II

Table 4 .
Energy cost data for case study II

Table 5 .
Energy consumption and carbon dioxide emission data for case study III

Table 6 .
Energy cost data for case study III

Table 7 .
Energy consumption and carbon dioxide emission data for case study IV

Table 8 .
Energy cost data for case study IV Downloaded by [] on [23/11/17].Published with permission by the ICE under the CC-BY license

Table 9 .
Energy consumption and carbon dioxide emission data for case study V

Table 10 .
Energy cost data for case study V

Table 11 .
EWI cost breakdown for two types of mid-terrace dwellings

Table 12 .
Basic payback cost analysis for retrofitted EWI at three mid-terrace dwellings Downloaded by [] on [23/11/17].Published with permission by the ICE under the CC-BY license Nevertheless, while the results Welsh Assembly Government, 2010a)d that Arbed is a really valuable method of funding retrofitted EWI, it is unlikely that it will contribute to meeting the requirements of the Climate Change Act 2008 and the Welsh government's own carbon dioxide emission reduction targets(Climate Change Act  2008, 2008;Welsh Assembly Government, 2010a).The authors cannot comment on the Arbed scheme as a whole; however, if similar results occurred as those presented in this paper, collectively the Arbed I scheme could have failed to meet its objectives related to the Climate Change Act 2008 and the Welsh government's carbon dioxide emissions reduction target (Climate Change Act